This blog is a great innovation. I hope it works as well as I think it will. Thank you for trying it!
In talking with many scientists about molecular manufacturing, I’ve found that they frequently over-generalize limitations. I hope it won’t sound too arrogant or too obvious if I list a few of the areas where scientists have tended to overestimate problems and thus limit their own problem-solving capacity:
1) Anything we or biology can do today establishes a lower bound, not an upper. For example, there is no fundamental reason to expect that chemical processes will be limited to the 10^-6 error rate of DNA transcription.
2) Recently a physicist asked me whether entropy wouldn’t build up in a nano-computational system and cause errors. My answer–that modest energy input could restore digital signals, and the resulting heat could be conducted away–satisfied him immediately, but without that conversation, he might have continued to assume that entropy would be a practical problem. Even theoretical limitations often arise from overgeneralized assumptions.
3) In fact, the word “entropy” deserves special honors. “Entropy” is a catch-all that turns into a muddle. Separate from my conversation with the physicist, an expert computer scientist has written that Babbage’s mechanical Analytical Engine could not have worked because entropy would build up and distort the signals–which is of course completely incorrect; simple detents can restore digital mechanical signals. Entropy always has a physical mechanism. Identify the mechanism, and you may find that it’s not as bad as you assumed, or that you’ve lumped together a practical problem with a much less troublesome theoretical limit–and the practical problem can be solved.
4) I’ve found that most nanotechnologists are used to thinking in terms of complicated and finely-tuned reaction conditions to achieve intricate and useful results. I don’t know if there is a theoretical or only a practical dividing line between the following two regimes: A) Complex conditions -> complex phenomena -> complex output. B) Simple programmable operations -> repeated many times -> intricate output. Computer scientists and engineers are comfortable with B). My experience suggests that most nanotechnologists are only comfortable with A).
I hope that some of these points are more useful than offensive. Thank you for considering them.